4 ms·

The first thing you learn when playing with genetic algorithms is that the mutation rate has a major influence on the outcome. Evolution doesn't work if the rat

by nom 10y ago

The first thing you learn when playing with genetic algorithms is that the mutation rate has a major influence on the outcome. Evolution doesn't work if the rate is too high or too low and within the working range it also shows a positive correlation between mutation rate and speed of evolution. It makes sense that the evolution of Tardigrades has slowed down tremendously by now.

p_zakharov 10y ago
As far as overall survivability for an animal is concerned, regardless of the mutation rate, one would be a little hard-pressed to imagine how they'd climb any further up whatever hill on the fitness landscape it is they're on.
- Ultimatt 10y ago
  Again just no. Tardigrades are predated by each other and plenty of other animals. Just because you can survive drought and radiation doesnt mean you can mechanically survive another organism evolved to eat you hunting you down.
Ultimatt 10y ago
No it doesn't make sense at all. The vast majority of natural evolution works through composability and shuffling of already evolved structures. So through sexual recombination. Many tardigrades can also selectively vegetatively reproduce fixating any mutation in the population. This means the rate of evolution has no reason to slow at all. It could even be quite fast as beneficial mutations have a mechanism to become widespread in the population straight away, unlike many other animals. Even more interesting is the level of possible horizontal gene transfer seen in the Tardigrade genome. So there is plenty of injection from other lineages to keep up diversity. If you want the genetic algorithms perspective on how that works check out Richard Watson's work https://mitpress.mit.edu/books/compositional-evolution https://mitpress.mit.edu/books/compositional-evolution But in general it's really quite foolish to imagine because you know something about GAs you have even a hint of what natural evolution is like. The two are very different. GAs are super simplistic and essentially only cover some basic concepts from 100yol evolutionary theory and mostly ignore essentially everything we know from modern molecular biology. Island GAs and Richard Watson's concepts of compositional evolution actually capture a lot of what happens. The reason is fairly simple if you put it in context of how molecular structures evolve, a good starting place is knowing what a protein domain is https://en.wikipedia.org/wiki/Protein_domain https://en.wikipedia.org/wiki/Protein_domain

by nom 10y ago

p_zakharov 10y ago

As far as overall survivability for an animal is concerned, regardless of the mutation rate, one would be a little hard-pressed to imagine how they'd climb any further up whatever hill on the fitness landscape it is they're on.

Ultimatt 10y ago
Again just no. Tardigrades are predated by each other and plenty of other animals. Just because you can survive drought and radiation doesnt mean you can mechanically survive another organism evolved to eat you hunting you down.

Ultimatt 10y ago

No it doesn't make sense at all. The vast majority of natural evolution works through composability and shuffling of already evolved structures. So through sexual recombination. Many tardigrades can also selectively vegetatively reproduce fixating any mutation in the population. This means the rate of evolution has no reason to slow at all. It could even be quite fast as beneficial mutations have a mechanism to become widespread in the population straight away, unlike many other animals. Even more interesting is the level of possible horizontal gene transfer seen in the Tardigrade genome. So there is plenty of injection from other lineages to keep up diversity. If you want the genetic algorithms perspective on how that works check out Richard Watson's work https://mitpress.mit.edu/books/compositional-evolution https://mitpress.mit.edu/books/compositional-evolution But in general it's really quite foolish to imagine because you know something about GAs you have even a hint of what natural evolution is like. The two are very different. GAs are super simplistic and essentially only cover some basic concepts from 100yol evolutionary theory and mostly ignore essentially everything we know from modern molecular biology. Island GAs and Richard Watson's concepts of compositional evolution actually capture a lot of what happens. The reason is fairly simple if you put it in context of how molecular structures evolve, a good starting place is knowing what a protein domain is https://en.wikipedia.org/wiki/Protein_domain https://en.wikipedia.org/wiki/Protein_domain